National University of Computer and Emerging Sciences

About: National University of Computer and Emerging Sciences is a education organization based out in Islamabad, Pakistan. It is known for research contribution in the topics: Computer science & The Internet. The organization has 1506 authors who have published 2438 publications receiving 26786 citations.

Classification of Skin Cancer Dermoscopy Images using Transfer Learning

Abstract: Cancer is the second death causing disease in the world. Skin cancer is a major type of cancer and increasing speedily over the past decades. The major causes of skin cancer are Ultraviolet radiation, smoking, DNA mutation and bad lifestyle. Skin cancer is diagnosed by employing clinical screening, dermoscopic analysis, histopathological examination and a biopsy. Early prognosis of skin cancer can help to cure cancer easily and It is a difficult task to diagnosis and classify skin cancer even for expert dermatologists at an early stage. Recent studies have shown that deep learning and transfer learning is very useful in the classification of skin cancer and medical diagnosis. In this paper, we propose an efficient way to classify skin cancer using deep learning. We have tuned the pre-trained MobileNet convolution neural network and trained over the HAM1000 skin lesion dataset. This transfer learning method has shown remarkable categorical accuracy, the weighted average of precision and recall 0.97, 0.90 and 0.91 respectively. This model is lightweight, fast and reliable. It will be helpful for dermatologists to prognosis the skin cancer at its early stage.

Design of normalized fractional SGD computing paradigm for recommender systems

Abstract: Fast and effective recommender systems are fundamental to fulfill the growing requirements of the e-commerce industry. The strength of matrix factorization procedure based on stochastic gradient descent (SGD) algorithm is exploited widely to solve the recommender system problem. Modern computing paradigms are designed by utilizing the concept of fractional gradient in standard SGD and outperform the standard counterpart. The performance of fractional SGD improves considerably by adaptively tuning the learning rate parameter. A nonlinear computing paradigm based on normalized version of fractional SGD is developed in this paper to investigate the adaptive behavior of learning rate with novel application to recommender systems. The accuracy of the proposed approach is verified through root mean square error metric by using different latent features, learning rates, fractional orders and datasets. The superiority of the designed method is validated through comparison with the state-of-the-art counterparts.

Effects of chemical reaction and magnetic field on a couple stress fluid over a non-linearly stretching sheet

Abstract: The boundary layer flow of an electrically conducting couple stress fluid over a non-linear stretching sheet is investigated in the presence of chemical reaction and magnetic field. The governing partial differential equations are reduced to coupled ordinary differential equations (ODEs) with the corresponding appropriate boundary conditions. The analytical solutions of the resulting ODEs have been obtained via the homotopy analysis method (HAM) in the form of series, and their convergence regions can also be found by the convergence control auxiliary parameter. The influence of pertinent parameters such as the magnetic, the couple stress and the chemical reaction parameters are discussed on the velocity and concentration profiles of the fluid. Hence, a comparison has been made with the previous results in the literature as a limiting case of the considered problem.

A review on low-complexity structures and algorithms for the correction of mismatch errors in time-interleaved ADCs

Abstract: In this paper we review the progress in the design of low-complexity digital correction structures and algorithms for time-interleaved ADCs over the last five years We devise a discrete-time model, state the design problem, and finally derive the algorithms and structures In particular, we discuss efficient algorithms to design time-varying correction filters as well as iterative structures utilizing polynomial based filters Finally, we give an outlook to future research questions